{"id":"W3006673017","doi":"10.1002/jrsm.1398","title":"Comparing machine and human reviewers to evaluate the risk of bias in randomized controlled trials","year":2020,"lang":"en","type":"article","venue":"Research Synthesis Methods","topic":"Meta-analysis and systematic reviews","field":"Decision Sciences","cited_by":28,"is_retracted":false,"has_abstract":true,"ca_institutions":"Capital District Health Authority; Institute of Health Economics; University of Alberta","funders":"Canadian Institutes of Health Research; Institute of Health Economics; Alberta Innovates - Health Solutions; Physiotherapy Foundation of Canada; Government of Alberta","keywords":"Blinding; Randomized controlled trial; Computer science; Medical physics; MEDLINE; Sample size determination; Medicine; Applied psychology; Psychology; Statistics; Surgery; Mathematics","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch","metaepi_broad","insufficient_payload"],"consensus_categories":["metaresearch"],"category_scores_codex":[0.9613858,0.0003113156,0.03045308,0.0009493275,0.0002630968,0.0006640845,0.002212001,0.00007225099,0.004014325],"category_scores_gemma":[0.9737784,0.00009884116,0.004915943,0.003085848,0.0003093047,0.00008994956,0.0004716413,0.0004823574,0.0001890478],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002975469,"about_ca_system_score_gemma":0.00008817553,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002420641,"about_ca_topic_score_gemma":0.00005014961,"domain_scores_codex":[0.07757422,0.8899349,0.02175007,0.001147778,0.009117885,0.0004751256],"domain_scores_gemma":[0.02095731,0.9649959,0.007262348,0.003833479,0.002439407,0.0005115821],"domain_codex":"methods","domain_gemma":"methods","domain_candidate":"methods","domain_consensus":"methods","study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0714725,0.0002954177,0.04068707,0.001313711,0.0127374,0.00001361429,0.01034255,0.001745631,0.01359938,0.01636753,0.02938227,0.802043],"study_design_scores_gemma":[0.2149473,0.0003861727,0.01297762,0.002094604,0.01225229,0.000005961582,0.006069827,0.6109159,0.007147806,0.06061711,0.0715026,0.001082767],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5023127,0.101016,0.2713241,0.04156169,0.0003316413,0.05103473,0.00005935137,0.0000229532,0.03233675],"genre_scores_gemma":[0.9018297,0.002045498,0.09380534,0.0001990704,0.0001002193,0.001375249,6.263056e-7,0.00002183748,0.0006225189],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8009602,"threshold_uncertainty_score":0.9968961,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.9609323859410629,"score_gpt":0.7227307998086823,"score_spread":0.2382015861323806,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}